A Critical View on the Lightning Protection International Standard

ProI. C. Bouquegneau
Pro-rector oI the Faculte Polytechnique de Mons, rue de Houdain, 9, B-7000 Mons (Belgium)
Christian.BouquegneauIpms.ac.be
Abstract. The Technical Committee TC81 (Lightning
Protection) oI the IEC (International Electrotechnical
Commission has Iinalised the new presentation oI its work
in Iive parts (IEC 62305-1 to -5) treating general
principles, risk management, physical damage, liIe
hazards, protection against electrical and electronic
systems within structures and some services entering the
structure. We criticise some options retained.
Keywords. Lightning Protection, Standardization.
1. INTRODUCTION
IEC standards are based on scientiIically proven theories
and technical experimentation world-wide taking into
account the international expertise in the matter. They lay
down requirements Ior the design and installation oI LPS
(Lightning Protection Systems) Ior structures and
buildings, the protection against lightning oI services
entering the buildings and the protection oI electrical and
electronic systems.
TC81 is achieving its Iirst cycle oI work when issuing
next year a complete standard (IEC 62305) in Iive parts
(IEC 62305-1 to 5, see the list in 2). The standard
provides the general principles to be Iollowed in the
protection against lightning oI a structure (including its
installations and contents as well as persons) and services
entering the structure.
The general principles are presented and criticised with
the latest contributions Iromworking groups oI IEC TC81
(and CLC TC81X, a similar European Committee inside
CENELEC), with comments Irom National Committees
concerned with lightning protection.
Direct and nearby cloud-to-ground discharges can be
hazardous to people, structures, their contents and
installations, as well as to services. Hence the application
oI lightning protection measures must be considered.
The need Ior protection, the economic beneIits oI
installing protection measures and the selection oI
adequate protection measures should be determined in the
terms oI risk management ; the risk management method
is reported in IEC 62305-2.
There are no devices or methods capable oI preventing
lightning discharges. Direct and nearby cloud-to-ground
discharges can be hazardous to structures, persons,
installations and other devices on them. The need Ior
protection and selection oI adequate protection measures
is determined in items oI risk assessment: National
Committees have generally interpreted and simpliIied the
international approach, which leads to some discrepancies.
Fortunately, a convenient Ilexible software will be oIIered
with the text oI the second part oI the standard: it is right
now on the testing stage.
The criteria Ior design, installation and maintenance oI
lightning protection measures are considered in three
separate groups:
- protection measures to reduce physical damages and liIe
hazards in a structure is reported in IEC 62305-3 ;
- protection measures to reduce Iailure oI electrical and
electronic systems (inside the structure to be protected) is
reported in IEC 62305-4 ;
- protection measures to reduce physical damages and
Iailure oI services entering the structure (mainly electrical
and telecommunication lines) is reported in IEC 62305-5.
In the new edition oI the standard there is no limitation oI
height oI the structures and buildings. Nevertheless,
railway systems and vehicles, ships, aircraIt and oIIshore
installations are still outside its scope.
The classiIication oI the structures depends on the
consequential eIIects oI lightning Ilash which can cause
damage to the structure, their contents or their
surroundings.
2
We are going to Iocus on the basic principles leading to a
not so straightIorward international consensus and
criticise some speciIic approaches, repelling some other
ones emphasised in some countries and essentially based
on more commercial than scientiIic arguments.
2. LIST OF IEC TC81 NEW STANDARDS
The actual list oI standards that is already or will be soon
issued by IEC TC81 is the Iollowing.
IEC 62305-1 Part 1: Protection of structures against
lightning : general principles:
it introduces terms and deIinitions, lightning current
parameters, damages due to lightning, protection needs
and measures, basic criteria Ior protection oI structures
and services as well as test parameters simulating the
eIIects oI lightning on LPS components;
IEC 62305-2 Part 2 : Risk management:
it introduces the risk assessment method, the assessment
oI risk components Ior structures and the assessment oI
risk components Ior services;
IEC 62305-3 Part 3 : Physical damage and life hazard:
it is related to lightning protection systems (LPS),
protection measures against injuries oI living beings due
to touch and step voltages and it oIIers a guideline Ior
design, installation, maintenance and inspection oI LPS ;
IEC 62305-4 : Part 4 : Electrical and electronic systems
within structures:
it considers the protection against Lightning Electro-
magnetic Pulses (LEMP): general principles; earthing and
bonding inside structures; magnetic shielding and line
routing, requirements oI surge protective devices (SPD),
protection oI equipment in existing structures;
IEC 62305-5 : Part 5 : Services:
telecommunication lines (Iibre optic lines and metallic
conductors lines), power lines, pipelines are concerned.
3. PRINCIPLES AND METHODS
3.1. Interception models
As was stated many times in the past, the conventional
lightning protection philosophy, the methods and their
practical implementation rest on a well Iound theoretical
and empirical basis with a vast experience Ior the
veriIication and validation oI the the so-called rolling
sphere method (RSM) associated to an electrogeometric
model (EGM) and based on a radius R equal to the
striking distance or Iinal jump distance; so Iar it is the best
model we have to work with in an international standard.
We personally regret that the standard introduces three
methods as iI they were diIIerent ones:
(1) the RSM (or EGM) method which should be so Iar the
only one to be recommended anywhere and at any time,
even iI it has not been evaluated yet on tall buildings; it is
a crude method which hides our insuIIicient understanding
oI the lightning attachment process but appears as the best
tool Ior the design and the positioning oI the air terminals;
in the Iuture we will improve this model taking into
account corners and edges oI very tall structures, but we
still need more Iield data about the striking oI such tall
structures; this model has already been somewhat
improved |1| and is now the reIerence in international
standards but it should also be reIined taking into account
both downward/upward leaders velocities and propagation
parameters; in this respect, we strongly support Gerard
Berger`s last approach as a new candidate to improve the
EGM model |2|;
(2) the protection angle method (PAM) only seems to be
maintained Ior historical reasons; in Iact it is simply
derived Irom the previous one (RSM) but is not a diIIerent
one; we think it should be suppressed or at least it should
not appear as a method diIIerent Irom the RSM; moreover
it is subject to limits oI air terminal height (indicated in
Table 2 oI IEC 62305-2, 5.2.2), because the protection
angle is not constant any more since it decreases iI the
height increases ;
(3) the mesh method (MM), sometimes (wrongly) called
Faraday cage method, can be too approximative iI several
metallic parts stand as high as and too close to the upper
mesh installed on Ilat planes or rooIs; anyway we think
that IEC documents should give a little more guidance
about the proposed height oI the mesh conductors above
the Ilat surIace they are supposed to protect.
Other methods have been proposed Irom time to time to
the international commissions, so Iar without success. For
example, the CVM method (CVM Collection Volume
Method) recommended by the Australian Committee has
received substantial criticism both concerning the method
itselI and the Iact that it was neither accepted nor analysed
by the international scientiIic community; Ior example,
the collection volume does not change when the leaders
downward/upward velocity ratio changes, which is not
realistic.
3.2. Air termination systems
The designer oI a lightning protection system should be
more aware oI the eIIiciency oI a horizontal conductor
with respect to a vertical rod when the tip oI this rod is
installed at the same height as the horizontal wire. The
beneIit oI vertical rods is generally and wrongly
emphasised compared to meshed conductors or catenary
wires.
Moreover a lot oI Iunny devices like 'aigrettes let people
Ialsely think that multiple tips improve the eIIectiveness
3
oI a single rod... though it is exactly the opposite (due to
the mixing up oI the space charges around the tips).
Side Ilashes must always be taken into account Ior
buildings and structures higher than 60 m, that is why it
seems suIIicient to install a regular lateral air termination
system (upper parts oI down conductors) on the upper part
oI such tall structures (generally on 20 oI the total
height), being careIul to the surrounding environment
(irregular shape oI the structure itselI or other separated
but neighbouring structures).
Ambitions and potential earnings involved in the design oI
more eIIective lightning receptors is an obvious
motivation Ior the invention and presentation oI a lot oI
diIIerent lightning protection systems and items, where
the claimed advantages have oIten been advertised,
unIortunately without veriIication oI their Iunctions and
validation oI their eIIects. So Iar parallel tests with simple
Franklin rods and various ESE (Early Streamer Emission)
devices exposed to natural lightning have shown no
signiIicant diIIerence in the attraction distance nor in the
number oI strokes to the diIIerent types oI rods. HopeIully
in the Iuture more eIIective lightning protection comp-
onents and systems could be developed but until such
systems are proven in a scientiIic sense their use should
not be allowed Ior objects where protection is required.
We have to remain reasonable and to be careIul when
issuing standards and guides. OI course IEC TC81
Iollowing conIirmed scientists does not advertise such
devices.
Nevertheless is it enough? Is it enough to ignore ESE
(French PDA,...), repellers, eliminators,... or just to say
that iI they are installed they have to be positioned as
conventional ones? The international standard looks to me
much too shy about the rejection oI these devices.
Radioactive air terminals are Iorbidden nowadays; they
had no preIerential interception eIIect but they were only
Iorbidden Ior their radioactive pollution.
Lightning repellers, dissipation array systems and other
eliminators can oI course not prevent the initiation oI
lightning in the thundercloud, nor avert any lightning
strike. ESE systems have never shown any superiority
over conventional systems and this should be emphasised
in all scientiIic conIerences on lightning protection, in
particular at this ICLP meeting.
3.3. Earth termination system
The earth termination system is a crucial part oI the light-
ning protection system, since it must disperse the lightning
current into the ground without danger to the people nor
damages to the installations inside the protected structure.
The transient behaviour oI earthed electrodes under high
peak impulse currents is a crucial point |3|.
Soil ionisation is still under consideration, it would
decrease the ground surge impedance only Ior very high
peak current values.
In any case a complete integrated earth termination system
with ring earth electrode (buried conducting loops or
Ioundation earth electrode) is always to be privileged to
decrease both the surge ground impedance and the
conventional earth resistance; moreover it is generally
better to install short-length vertical or inclined electrodes
in multiple ground electrode arrangements (prismatic or
pyramid-truncated arrangements) than single deep ground
electrodes; oI course each arrangement must be connected
to a down conductor.
Generally, the ring earth electrode is installed with addi-
tional radial, vertical or inclined short-length electrodes.
The author showed |3| that the optimal conIiguration shall
stand with multiple earth electrode arrangements with 3 or
4 short-length inclined electrodes separated by about 2 m
Irom each other and making an angle oI 30 with respect
to the vertical direction.
3.4. Lightning current parameters
Parameters oI lightning currents are selected Irom CIGRE
(Conseil International des Grands Reseaux Electriques a
Haute Tension) ; lightning current peak values and
waveshapes result Irom various classiIications : short and
long duration components, leader polarity and direction,.
Let us notice that the so-called M-components are not
considered in the standard |4|; we should say a word on
them, at least to mention that the protection is covered
when considering the other types oI waveshape currents.
We need more and more international data to work with a
reliable statistics oI lightning peak currents and other
parameters on general structures and buildings, because
CIGRE data were essentially brought Irom transmission
lines and tall structures.
Besides the peak value oI the Iirst stroke current, the
important parameters are the maximum rate oI rise
(induced overvoltages and dangerous sparking), the Ilash
duration and its total charge (thermal eIIects) and the Ilash
speciIic energy (selection oI metallic conductors Ior the
LPS and earthing system).
A better evaluation oI lightning current parameters and
particularly their height dependency taking into account
downward and upward Ilashes should be perIormed by the
international scientiIic community.
4
3.5. Lightning protection levels
Four lightning protection levels LPL (I to IV, with Iour
types oI relevant protection measures Ior the design oI
LPS) are introduced. For each one, a set oI maximum and
minimum lightning current parameters is Iixed.
The maximum values oI lightning current parameters
relevant to LPL I will not be exceeded with a probability
oI 99 ; they are reduced to 75 Ior LPL II and to 50
Ior LPL III and IV. The minimum values oI lightning
current amplitude Ior the diIIerent LPL are used to derive
the rolling sphere radius R (RSM or EGM method) in
order to deIine the lightning protection zone which cannot
be reached by direct strikes, a minimum peak current oI,
respectively, 3 kA (LPL I), 5 kA (LPL II), 10 kA (LPL
III) and 16 kA (LPL IV) leads to respective values oI the
rolling sphere radius R equal to 20, 30, 45 and 60 m. They
are used Ior the positioning oI air terminations in the
external protection and to deIine the lightning protection
zone LPZ 0
B
(protected against direct lightning strikes) in
the internal protection.
An important question can be raised. Why do we limit
level I to a probability oI 99 , since we are sometimes
obliged to take into account complementary measures to
come to a level I ? Level I should surely be the highest
level oI protection, why not 100 even iI it looks
unrealistic? With 99 , we are in some uncomIortable
position in Iront oI the protection oI very dangerous
structures (inIlammable and explosive environments).
4. RISK MANAGEMENT
The protection measures must be applied taking into
account the risk management method which is reported
in IEC 62305-2, this method provides a procedure Ior the
evaluation oI the total risk to be compared with an upper
limit oI tolerable risk; this procedure allows the selection
oI appropriate protection measures to be adopted to reduce
the risk below a tolerable limit.
In the international standard the deIinition oI risk R is
peculiar: it is the probability oI having an annual loss in a
structure or its content. For the purposes oI this standard,
Iour types oI loss (L
1
to L
4
) are considered and each one
corresponds to a relevant risk :
R
1
risk oI loss oI human liIe,
R
2
risk oI loss oI service to the public,
R
3
risk oI loss oI cultural heritage,
R
4
risk oI loss oI economical value;
but each typical risk (R
1
to R
4
) is also the sum oI diIIerent
components R
X
(X A, B, C,., i.e. touch and step
voltages, dangerous sparking and thermal eIIects, over-
voltages,... ); each risk component R
X
depends on the
point oI strike and on the annual number oI dangerous
events N
X
attached to X, the related probability oI damage
P
X
(damage to the structure) and the consequent annual
loss L
X
due to a single lightning Ilash (related to the total
amount oI persons or goods) so that
R
X
1 exp(- N
X
P
X
L
X
) N
X
P
X
L
X
.
The various risk components are analysed and summed up
to end up with a total risk R :
R N P L ;
This total (level oI) risk (Irequency oI annual loss in a
structure due to lightning) is then deIined as the probable
annual loss in a structure due to lightning.
We must insist on some arbitrariness in the evaluation oI
the diIIerent components oI this calculation, particularly
the assumed values oI the probabilities oI damages which
are most oIten rounded oII to negative powers oI 10 or
integers (1 to 6, and 9) multiplying them; unIortunately
we can never be more precise in this matter but we can be
in Iront oI very unprecise values oI the total risk! We must
suppose that people who consider the Iinal result oI the
calculation (or the result shown on the screen by applying
the proposed soItware) are clever enough to interpret the
Iaint diIIerences when applying this method!
This total risk R must be compared to the tolerable value
oI the risk R
T
and always remain smaller or equal to R
T
;
this condition has to be satisIied Ior each type oI damage.
R
T
is a little arbitrarily deIined (under the responsibility oI
national body concerned!) Ior the Iirst three types oI
possible losses (losses oI social value), whose suggested
typical values are :
10
-5
Ior the loss oI human liIe,
10
-3
Ior the loss oI service to the public
10
-3
Ior the loss oI cultural heritage.
For the loss oI economic value (L
4
) a private decision will
be taken by the owner or the designer oI the structure
under their own reponsibility.
What is the precise meaning oI a suggested 10
-5
tolerable
risk Ior the loss oI human liIe? Why not 10
-4
or better 10
-6
or even better...? Does human liIe have a price? This is
another crucial point. Each choice could be justiIied by
some people and denied by other people... That is why
the values oI these tolerable risks are under the
responsibility oI the various National Committees, though
the economic tolerable risk is simply leIt to the owner oI
the structure or to the designer oI the LPS.
We would also like to underline the tremendous work
done by ProI. C. Mazzetti and Z. Flisowski in the IEC
TC81 WG devoted to the risk management problems and
we take this opportunity to warmly congratulate them |5|.
5
In risk assessment, it is crucial to know the average value
oI the regional lightning Ilash density (N
g
). We have to be
careIul in the interpretation oI diIIerent phenomena: the
Ilash density is not the stroke density but people
sometimes mix up both concepts. In temperate regions,
the stroke density is about 4 times the Ilash density;
moreover there is only one Ilash to ground Ior
approximately 3 inter/intra-cloud Ilashes. Fortunately,
modern lightning location systems help to discriminate
between all types oI discharges.
5. BASIC CRITERIA OF PROTECTION
A large consensus showed up in the elaboration oI the
basic criteria of protection :
1) protection against phvsical damages (Iire, explosion
danger and liIe hazards; see Part 3 oI the standard, i.e. IEC
62305-3) with an eIIicient LPS both
- external (interception, electric current conducted to
earth, dispersion into earth) and
- internal (preventing dangerous sparking within the
structure by equipotential bonding and separation
distances);
2) protection against LEMP based on the principle oI
(outer and inner) lightning protection zones (characterised
by signiIicant changes oI the LEMP severities compatible
with the immunity level oI the internal systems) with
protection measures which are essentially earthing,
shielding, bonding and line routing (see Part 4 oI the
standard, i.e. IEC 62305-4); a Iull LPM system (capable to
reduce the risk oI permanent Iailures oI electrical and
electronic systems which are sensitive to energies as low
as some mJ) will protect against conducted surges as well
as against radiated magnetic Iields; an essential point is
the coordination oI a set oI surge protective devices (SPD
system) which must be designed in liaison with the
standards issued by other technical committees (TC64,
SC37A,...); an important installation problem is the
selection oI the 'protection distance, i.e. the maximum
distance along the circuit at which the equipment will still
be protected: some proper relations are given in IEC
62305-4, but the user will need to evaluate these
protection distances by computer simulation;
3) protection of services entering the structure (cables,
telecommunication lines,.; see Part 5 oI the standard, i.e.
IEC 62305-5); unIortunately the work is still in progress
in this part oI the standard and we are waiting Ior experts
in pipelines and energy lines to work out on the missing
meaures oI protection; the telecommunication lines using
metallic conductors or optical Iibre cables are already
considered.
6. CONCLUSION
IEC TC81 has arrived at a substantial work on the means
oI lightning protection.
Not surprisingly, Europe adopts most oI the IEC TC81
standards inside CENELEC (CLC TC81X) ; the european
and the international commissions Iollowed the same
procedures with parallel voting inside the various National
committees.
Though the work is surely not perIect yet, we are entering
the maintenance period which should be used to improve
the standard. Some hints are proposed in this paper.
Anyway all the National committees should adopt this
international standard on lightning protection avoiding to
promote Iancy devices which do not comply with it.
REFERENCES
|0| See the list oI Iive IEC TC81 parts oI th international
standard in the text ( 2): IEC 62305-1 to -5.
|1| CIGRE Task Force 33.01.03 (Bernardi M., Bondiou-
Clergerie A., Cooray V., Dellera L., Pedersen Aa.,
Rhling F.), 'Lightning Exposure oI Structures and
Interception EIIiciency oI Air Terminals, CIGRE, 1997.
|2| S. Ait-Amar and G. Berger, 'A 3-D numerical model
oI negative lightning leader interception. Applications to
the collection volume construction, ICLP2004, Avignon,
Sept. 2004.
|3| C. Bouquegneau and B. Jacquet, 'How to improve the
lightning protection by reducing the ground impedances,
ICLP, the Hague (NL), Sept. 1983, paper 2.3, p. 89-96.
|4| Vl.A. Rakov and M.A. Uman, 'Lightning: Physics and
EIIects, Cambridge University Press, 2003.
|5| C. Mazzetti, M. Pompili and Z. Flisowski, 'Review oI
application problems oI probabilistic approach to the
lightning hazard assessment, ICLP, Florence, Sept. 1996,
paper 7.1, pp. 707-712.